Within the European Union, the call to avoid tax abuse has been loud for over a decade. The European Parliament’s FISC Committee, previously chaired by Paul Tang, already voiced that the EU should take action against the use of shell entities and letterbox companies.
Unshell Directive
In 2021, the European Commission proposed the Unshell Directive, still incorrectly being referred to as ATAD3. The objective of the Unshell proposal was to prevent “practices of tax avoidance and evasion linked to the use of undertakings which do not perform an economic activity, even if presumably they are engaged with economic activity and therefore do not have any or have only minimal substance for tax purposes” (preamble, point 3). The proposal took a broad approach, funneling all EU resident companies until only the ‘real shell entities remained. Therefore, the starting point was that all companies were in scope of the proposal. However, they could be out-scoped based on a seven step funneling approach. Examples for being out-scoped included situations where the companies were EU regulated entities, they were lacking tax motives or they had sufficient substance. The substance requirements included three elements: (i) the undertaking has own premises in the Member State, or premises for its exclusive use; (ii) the undertaking has at least one own and active bank account in the EU; and, (iii) either one or more of the directors of the majority of the full-time equivalent employees meet certain criteria establishing a strong connection to the Member State of residence of the company concerned. If, in the end, the company concerned was considered a shell entity, the entity itself would not be entitled to a tax residence certificate anymore, and the jurisdictions of the investor in the shell entity and the investment of the shell entity should introduce tax consequences, e.g. by treating the shell entity as a transparent entity.
The proposal for an Unshell Directive faced criticism from the start. Mainly the seven step system to determine whether a company qualifies as a shell entity was heavily criticized, as well as the mandatory tax consequences attached to that qualification. After all, it would only be EU entities that would be forced to apply those tax consequences. Non-EU investor or investment jurisdictions cannot be forced under EU law to address tax consequences via an EU Directive.
Even though the European Parliament suggested several amendments to the European Commission’s proposal, the European Council already had indicated not to be in favor of the Unshell proposal, especially because of the substance criteria[1] and the tax consequences[2]. Even though several EU Presidencies wanted to reanimate the proposal, for instance by first introducing exchange of information on the use of shell entities and only at a later stage add tax consequences to the use of shell entities,[3] the Council could not find unanimity. In October 2025, the European Commission announced to withdraw the proposal for an Unshell Directive.
Substance as a DAC6 hallmark?
The fact that the proposal for an Unshell Directive would be withdrawn, did not mean the end of the discussion on the use of entities with little substance. As the original proposal already consisted of two components – i.e., exchange of information and tax consequences – , the European Commission suggested to include substance requirements in the Directive on Administrative Cooperation. This Directive regulates the automatic exchange of information within the EU. Its 5th amendment, known as DAC6, determines that intermediaries or – ultimately – taxpayers need to report reportable cross-border arrangements within 30 days after the advise is made available for implementation, ready for implementation or the first step of implementation has been made.
The suggestion to add substance requirements as a new hallmark striked me. I doubt whether anyone would suggest to use an entity with too little substance. As the advice would be the triggering event, probably no reportings would be made. If, later, it appears that factually an entity would not have sufficient substance, that would not be incorrect for DAC6 purposes, as on paper the advice concerned the use of an entity with sufficient substance. This, by the way, even still apart from the discussion what sufficient substance would need to be.
DAC Recast and substance requirements
Apparently, the European Commission found a link to introduce substance criteria in the DAC Recast. Essentially, the aim of the proposal for a DAC Recast is to have a new starting point for exchange of information after the 2011 DAC has been amended for in eight substantive occasions. Besides that, the European Commission’s desire to simplify existing legislation will be taken on board in the revisioning process.
As a part of the recast, the inclusion of substance requirements was reassessed. The European Commission acknowledged that DAC6 hallmark D2 already includes certain substance criteria that should be further developed. However, it also understood that including exact substance criteria in the proposal for the DAC Recast would slow down the adoption process of the directive as Member States might, or will, oppose the exact wording of the substance criteria. A lesson learned from the Unshell proposal. Therefore, as a solution, the proposal for the DAC Recast suggests to develop common rules to define economic substance via a Council implementing act. The effect, aspired by the European Commission, would be that the proposal for the DAC Recast could be adopted before year end. It would also buy them time to develop and adopted the actual substance criteria separately at a later stage and through a swifter procedure.[4]
Council implementation act
The question that can be raised is whether this approach is correct. Even though it can politically easily be explained why the discussions on the definition of sufficient substance is deferred, the chosen instrument to later actually add the substance criteria is debatable.
Harmonization in the field of direct taxation should be based on Article 115 TFEU, requiring unanimity in the Council and prescribing the directive to be the harmonizing instrument. However, if uniform conditions for implementing those directives are needed, it is possible to confer implementing powers to the European Commission in the basic directive. One could, in this respect, think of the creation of standard forms. By granting competences to the European Commission, the basic legislation would not need to be amended and practical aspects can be arranged. The legal basis for the use of implementing acts can be found in Article 291 TFEU. However, only in duly justified specific cases and in the cases provided for in Articles 24 and 26 TEU, implementing powers can be conferred to the Council. The Articles 24 and 26 TEU relate to the common foreign and security policy and can, therefore, not sort effect. Only if substance requirements, or taxation in general, can be considered a ‘duly justified specific case’, the competence on the implementing act could be conferred on the Council instead of the European Commission.
Earlier, the CJEU judged that Article 291 TFEU, whereby the Council may reserve the implementing power to itself ‘in duly justified specific cases’, requires that the Council should state in detail the grounds for the decision to reserve the implementing powers to itself. In relation to the legal predecessors of Article 291 TFEU, the Court had already stated that the Council must properly explain, by reference to the nature and content of the basic instrument to be implemented or amended, why exception is being made to the rule that the Commission has the delegated implementing power. Even though the preamble to the proposal for the DAC Recast does provide clarity on the creation of the delegation power, it only explains to a limited extent why this competence for the implementing act should be on the Council instead of on the Commission, as would be the main rule. The preamble specifies in point 105: Substance requirements “are expected to have an impact on Member States’ executive and enforcement powers in the field of direct taxation, as well as on Member States’ tax bases”. To my opinion, the fact that harmonization in the field of direct taxation would require unanimity would, in itself, not be sufficient to be a ‘duly justified specific case’. Besides that, the DAC Recast would only relate to exchange of information and does not per se have a clear and direct impact on the Member States’ tax bases. As such, I feel the instrument of an implementation act should not be the way to go.
Article 291 TFEU is further developed in Regulation 182/2011. The Regulation creates the procedures under which implementing acts can be adopted. It is silent on how ‘duly justified specific cases’ can be defined. However, in the field of taxation, in principle, only the so-called examination procedure for implementing acts can be followed. In relation to the substance criteria, the examination procedure to be followed would mean that the European Commission proposes the substance criteria. These need to be reviewed by a Committee of Experts of the Member States. If they approve, the European Commission adopts the implementing act.
The Regulation does allow the Council and the European Parliament to be informed on the status of the implementation act. If the case cannot qualify as a ‘duly justified specific case’, the Council cannot decide on the implementing act. And, even though Article 11 of Regulation 182/2011 does allow the Council and Parliament to block drafts at any stage if they think the European Commission exceeds its competence, that option only applies when the ‘normal legislative procedure’ is followed in relation to the basic directive. For harmonization in the field of taxation, that is, of course, not the case.
Is this a good idea?
The reason for an alternative route and excluding substance criteria from the original proposal for the DAC Recast seems a good idea in order not to lose valuable time. I agree with the European Commission that defining substance criteria at this stage would delay the adoption of the proposal as a whole.
It is, however, the alternative route of using a Council implementing act that would not have my preference. It leaves the creation of the actual substance criteria to the European Commission, without any consent to be given by the Member States or the European parliament. It is still questionable whether the Council could actually be the institution adopting the implementing act, as – at least for now – it lacks a sufficient motivation on why there is a duly justified specific case following which the Council is more competent to adopt the implementation act. And, at a later stage, national parliaments of the EU Member States also have nothing to say anymore on the substance criteria. Therefore, too much power is vested in the European Commission at this point.
To me, that would be a bridge too far, especially since the way forward is only opted for in the sake of speed. Speed to adopt the proposal for a DAC Recast, and speed to – at a later stage – try to adopt substance criteria in a quicker and easier process. From a governance perspective, there are a lot of issues in relation to this.
Also substantively, several issues can be pinpointed. Of course, one could say that in the case of the DAC Recast, it only concerns the definition of ‘substance’ for the purpose of the exchange of information. The qualification as an entity without sufficient substance would not immediately have tax consequences. It can, however, not be excluded that at a later stage the discussion on the prevention of the use of shell entities resurfaces from a more substantive tax perspective as well, for instance by calling for disallowing tax benefits for the use of companies without sufficient substance. We run the risk that, in that case, the first definition on ‘substance’ to look to would be the one in the implementing act. If that would be too broad, the indirect effect of the creation of the substance definition in the currently proposed implementing act would be enormous. Even though an entirely new definition could then be established for substantive tax purposes, I would fear that the European institutions do not again want a decluttering and simplification assignment, by streamlining the substance requirements in the DAC Recast implementation act and the -then – newly introduced substantive EU directive.
Are there no alternatives to prevent the use of shell entities? Yes, there are. One could, for instance, solve it ‘the royal way’. Remove the reference to the implementation act from the proposal, adopt the proposal and immediately start the negotiations on the first amendment to the new directive. The procedure to be followed would be the special legislative procedure under Article 115 TFEU, safeguarding all legal and institutional aspects, including the roles of Council, parliaments and other stakeholders. Or, really thinking out-of-the-box, one could consider to use substance criteria as rebuttable assumption of (non-)abuse for the ATAD GAAR.[5] If a company has sufficient substance, it would not be considered abusive unless the tax authorities prove otherwise, and vice versa. This would, however, to my opinion only work if no ‘one size fits all’ solution is opted for by the Member States. Sufficient (relevant) substance should be an open norm in this respect. What would be sufficient substance for one type of company would not necessarily be sufficient relevant substance for the other. This would also be a relevant distinction for the CJEU’s anti-abuse case law (C-504/16 and C-613/16), indicating that all relevant facts and circumstances need to be taken into account for determining whether a case is abusive. Applying too rigid criteria for all companies appear not to meet the standard.
[1] E. Lamer, EU to Discuss Restoring Some Unshell Substance Criteria, TNI 2023-1944.
[2] E. Lamer, EU Countries to Discuss Limiting Tax Fallout for Shell Companies, TNI 2023-4720.
[3] E. Lamer, EU Council Presidency Suggests a Two-Step Approach on Unshell, TNI 2023-28561.
[4] E. Lamer, DAC Recast Draft Defers Debate Over Economic Substance Criteria, 2026 TNTI 115-2.
[5] Jasper Korving, Concurrence of EU Direct Tax Directives: Hierarchy of Norms, Interpretational Solutions, or Decluttering?, Intertax 2026/4, paragraph 3.4.